Journal Logo

Special Article

The Case of the Inadvertently Triggered Laser

An Historical Example of Simulation-Enhanced Adverse Event Investigation

Cooper, Jeffrey B. PhD

Author Information
Simulation in Healthcare: The Journal of the Society for Simulation in Healthcare: June 2021 - Volume 16 - Issue 3 - p 185-189
doi: 10.1097/SIH.0000000000000483
  • Free

Abstract

Although simulation is now widely applied for healthcare education and patient safety, its use for investigating adverse events has not been so widely adopted. There are numerous techniques for accident investigation1; root cause analysis (RCA) is the most commonly used in healthcare and is now a mainstream instrument for patient safety.2 It is used for learning how an adverse event evolved and applying that knowledge to actions that will reduce or eliminate recurrence. A recent enhancement of the RCA concept, RCA and Action (RCA2), more strongly pursues actions aimed at reducing the same or similar events. There are different approaches and tools for conducting an RCA. Although simulation is one of those, it is still not in common use, although there are numerous studies of its efficacy and suggestions for its use.3–6 An historical, previously unpublished, use of simulation is described here as an illustration of how a systematic re-enactment of an adverse event can reveal critical, actionable, root causes that otherwise likely would not be uncovered.

THE CASE

In the mid-1980s, a late-teenage female patient was undergoing a laser excision of a benign genital lesion. The patient was in the lithotomy position, with cotton surgical drapes over both legs, loose weave cotton surgical towels around the groin operative site, and several layers of the cotton surgical towels covering the surgical table under the patient's buttocks. The surgeon was using a CO2 laser (Biolas 40; model 40S; Advanced Biomedical Instruments, mfg c. 1983). Toward the end of the short procedure, the surgeon placed the laser wand, which was mounted on an articulating arm, slightly above the patient's left thigh, while continuing the procedure with other instruments. Shortly after, one of the operating team noticed an odor of smoke, which was then visibly seen emerging from around the drapes. The drapes were quickly removed, and active flames were observed coming from the cotton covering on the table. The flames were doused with saline, but not before the patient had suffered third degree burns on the inferior aspect of the left thigh and calf.

Immediate Follow-up

A surgeon specializing in treating burns was called and became engaged in short- and long-term treatment of the injuries. A biomedical technician, responding to the call for assistance, entered the room and sequestered the laser equipment, including the drapes and instruments used in the case.

A senior biomedical engineering leader (this author), having been contacted by operating room biomedical engineering staff, also came to the scene to supervise use of a then newly devised set of guidelines for actions after an anesthesia-related adverse event. (While this event was not anesthesia-related, the protocol still generally applied).7 He first ensured that immediate care of the patient was proceeding and that the sequestered equipment was secure from possible use or tampering. At the request of the operating surgeon, he and the surgeon conferred about how to approach the patient and a waiting family member. With consultation from the hospital risk manager, the patient's family member was informed of the injury and only the clearly known aspects of what had occurred, although the surgeon suggested that a failure of the laser may have been the cause.

INVESTIGATION

The immediate cause of the fire was presumed by those present to be due to some unknown mechanical or electronic failure that caused the laser to trigger spontaneously and ignite the drapes. The manufacturer of the equipment was no longer in business. There were no service nor other personnel to confer with other than the local biomedical engineering group that serviced all operating room equipment. The equipment intentionally was not tested immediately by the local biomedical engineering group to ensure that there would be no suspicion of tampering before inspection by an outside, impartial entity.

With permission and financial support from the captive insurance company that insures the hospital and its physicians, a hospital-fire incident investigation expert (Mark Bruley, ECRI Institute, a nonprofit agency, Plymouth Meeting, PA) was engaged to assist in investigating the cause of the event. The investigator and an assistant with experience in healthcare accident investigation arrived at the hospital the day after the event with the objective of ascertaining the cause of the fire and injuries to the patient. To do that, he, with assistance from this author, led a re-enactment of the event, what would now be called a “simulation.” The intent was to recreate the operating conditions with as much fidelity as practical to enable re-enactment of the process of the events in search of evidence for establishing the probable cause(s).

The information presented here is based on recollections of the author, the accident investigator, a report to the hospital of the investigator's findings, and a portion of the video that was made during recreation of the event (only a 5-minute edited version is extant).

The investigation was led by the accident investigator. We were permitted to use a vacant operating room in which we gathered a surgical table, an instrument-stand similar to that used in the procedure, the laser and its footswitch, and a Resusci Anne CPR mannequin used for training in the hospital. The surgeon, scrub, and circulating nurses who were present during the procedure were present during the simulation. The report of the event does not state specifically which staff other than the surgeon were present during the simulation, only that there were “other staff.” The mannequin was placed on the table with the legs draped in the lithotomy position. The drapes used in the case, some of which were partially burned, were put in the position as they were in the case according to the recollection of the nurses. The laser and its footswitch (Fig. 1) were placed where they had been during the case. The investigator asked the surgeon to describe and reproduce the steps taken during the operation, with attention to how the laser was used, the placement of the articulating arm, and use of the foot switch.

F1
FIGURE 1:
Footswitch used in the case.

The surgical towels and drapes immediately around the operative site where wetted with sterile saline. During the procedure, the surgeon used the laser scalpel for cutting and once for hemostasis. The laser was set at 13 W and held approximately 10 cm from the target tissue during coagulation, which lasted approximately 10 seconds. After controlling the bleeding, the surgeon placed the handpiece against the medial aspect of the patient's left thigh as a temporary resting place until further use. As was her routine, using her right foot, the surgeon then pushed the foot switch out of the way. She described this as a preventive measure, to avoid accidentally hitting the foot switch. It was noted that the surgeon was wearing relatively heavy clogs.

The 4 sequestered drapes were examined. The bottom drape, which was under the buttocks of the patient during the procedure, had a small burn hole and some blood stains. A loose weave towel that was placed on the bottom drape had a large burned area. A drape above that had a small burn hole. When the drape was placed by the scrub nurse the way it had been in the procedure, the small burn hole lined up with the large burn area in the drape below it. The outermost drape placed over the legs had a large burn area. The scrub nurse was unable to position that drape precisely relative to the patient because of the size of the burned area and uncertainty as to the angle of placement on the patients left leg. However, the burned area generally corresponded to the area where the laser handpiece rested. There was no corresponding damage visible on the operating table or its mattress.

Examination of the laser revealed that it was not equipped with a standby operating mode. In addition, the laser did not have an activation tone that would sound when the laser was firing.

Under supervision of the incident investigator, an operating room biomedical engineering technician repeatedly test fired the laser pointed at a wooden tongue depressor. The laser fired appropriately making a burn mark on the tongue depressor at the location of the visible aiming beam. There was no noticeable sticking of the footswitch or spontaneous firing of the laser.

It was noted that the laser uses a jet of nitrogen at the headpiece tip to clear away smoke. This jet would be capable of extinguishing burning cloth close to the tip, but it can be blocked from affecting lower layers by an overlying layer of cloth. We demonstrated this phenomenon by test firing the laser at 13 W (the setting used during the surgery), into several different types of cloth that were used during the procedure. The upper drapes were a tightly woven, high thread-count, cotton fabric. They did not ignite easily. The loose weave absorbent cotton towels that were immediately adjacent to the operating table ignited easily.

The laser foot switch was opened and examined. It had 2 internal switches and a protective cover over the pedal to help prevent inadvertent activation. When the pedal is depressed, the switches start the gas flow and open the shutter of the beam path. The thick metal cover around the switch prevents activation from above and the sides. It was noted that the foot switch required very little force for activation of the laser as compared with other foot switches familiar to the investigator. No exact measure of the force was made.

PRESUMPTIVE CAUSE OF THE FIRE AND INJURIES

Although we did not completely recreate the exact conditions during the procedure nor attempt to cause a similar fire with fresh drapes, the information gathered led to a presumptive explanation of causality by the incident investigation expert. When the laser wand was moved aside and placed over the drapes when not intended to be used, the surgeon unintentionally activated the switch when intending to move it out of a path of potential accidental activation. Because of her heavy clogs and because the switch was more easily activated than it should have been, she did not feel the switch being depressed. The laser, which had no standby mode and did not make a sound when firing, momentarily fired. The beam went through the top 2 drapes and hit the dry, cotton towel underneath. The jet of nitrogen cooled the uppermost drape, preventing it from igniting, which it likely would not have in any event because it was relatively difficult to ignite. The cotton towel ignited and burned slowly at first, creating an increasing hotter environment under the drapes, which trapped the heat and gas.

That trapping likely prevented earlier detection of the smoke and flame. The upper most drapes then ignited from the flame and heat below.

The investigator offered 5 suggestions for actions to lessen the likelihood of occurrence. Given the age of the laser, the demise of the manufacturer and cost and liability associated with modifications to it, the laser was retired from service. It was replaced with a more modern laser that included additional safety features that would have prevented this event, especially inclusion of a standby mode by which the laser is not turned on until needed by the surgeon. We do not recall with certainty if other suggested changes, for example, deployment of a laser nurse or technician whose sole responsibility is to monitor the laser when it is in use or using less flammable drape material, were adopted immediately after this event.

DISCUSSION

The investigation of this tragic case provides several lessons including (1) usefulness of a procedural guideline for actions after an adverse healthcare event7; (2) importance of sequestration of all materials and equipment after an adverse event; and (3) the efficacy of simulating (recreating) realistic conditions of an adverse event to learn immediate and some underlying causes of the event. We did not use the more detailed, thorough, formal RCA procedures that are now expected when investigating events such as this. However, the methodical investigation technique and the use of simulation, especially with the actual personnel involved, were likely pivotal to revealing the necessary elements that led to the patient's injury. Without such a re-enactment/simulation, we think that it is unlikely that the proximal cause could have been ascertained.

The focus of this discussion is on the use of simulation for investigation of healthcare adverse events to determine causality and prevention. This event occurred about 30 years ago, before the emergence of modern simulation in healthcare for any purpose and before the introduction of RCA and other adverse event investigation techniques as risk management tools. Thus, we did not use RCA or other tools such as fishbone diagramming, 5 whys, or other techniques that might be used were a similar event to happen today. However, there are aspects of this investigation that are exemplars of effective use of simulation to ascertain elements of causality that can inform design strategies and tactics for preventing similar events. In addition, there are lessons to be learned from aspects of the investigation that were suboptimal.

What Went Well

The simulation was pivotal to identifying several root causes and the most probable path leading to the patient's injury. Critical to that was the sequestration of the equipment, especially the drapes, which in many situations would have been discarded during the immediate response to the fire and caring for the patient. The biomedical technician who sequestered the involved items did so without the benefit of any training or protocol that directed such action. Being fairly senior and experienced, he likely had been involved in other events from which he may have learned the importance of maintaining the evidence for use in investigation of the event. This author has personally been involved in investigating or reviewing other events in which critical evidence was not sequestered, thwarting an informed process to ascertain causality.

The agreement of the insurance company to engage an incident investigation expert, particularly on short notice, aided in timely investigation. The cooperation of the operating room leadership and hospital administration to make personnel and a room available was also vital to a successful investigation. The expertise and experience of the incident investigator were likely critical to a successful investigation. Despite this author's experience in accident analysis, I would not have been as methodical in conducting a surgical fire incident investigation. Specifically, the skills of an experienced investigator are needed to take sufficient care in setting up the scene, use video recording effectively (which was not required but would have been useful in establishing proof for legal or other action), and especially to consider the various ways in which fires are started and propagated.

What Could Have Been Done Better

Although we are fairly certain of the key elements of this event, further recreation of the process of igniting the drapes might have been useful. That could not have been done safely in any setting but one for that purpose, for example, a forensic or accident investigation laboratory. The exact setting could have been recreated with fresh drapes and sheets of the same types used and attempts made to start a fire using short bursts from the same laser. Such an experiment would have been costly, including shipping of the laser, travel for personnel. However, if there had later been any skepticism, for instance, from lawyers seeking to place blame or from the surgeon accepting the explanation in the absence of such evidence, a more definitive experiment would have established complete certainty. As it was, the evidence gathered was sufficient to convince those involved in follow-up that this causal chain was plausible and that the remediation steps were required.

The footswitch likely played a more important role in this event than is suggested in the investigator's report. Sometime later, we disassembled the switch to assess its mode of action its mode of action, which is a simple spring mechanism that causes an electrical contact to be made and broken. The force required to activate the footswitch seemed light relative to what it could be to make accidental activation less likely. We were unaware of standards or specifications for surgical laser footswitches, and the force was not measured during the initial investigation or during our disassembly. Later, standard development for footswitch force activation for an analog device—electrosurgical units—required a minimum activation force of 10 N (approximately 2.25 lb; ANSI/AAMl/l E G6 0601-2-2:2009. Medical electrical equipment—part 2–2: requirements for the basic safety and essential performance of high-frequency surgery equipment and high-frequency surgical accessories).

Use of Simulation in Healthcare Adverse Event Investigation

Simulation is widely used in other high-risk industries for investigation of the causes of adverse events and accidents. A classic illustration is in the recreation of the final moments of Eastern Airlines Flight 401, which crashed in the Florida Everglades in 1978 (NTSB. Aircraft Accident Report NTSB-AAR-73; 1973). A documentary, “Why Planes Crash,” demonstrates how re-enactments, especially when guided by the cockpit voice recording, are critical to uncovering causality (https://www.youtube.com/watch?v=ibtmrSsk1NU).

There are several reports of the use of simulation of various kinds in conducting an investigation after a healthcare adverse event. In an experiment to evaluate the potential value of simulation for revealing nontechnical causes of adverse events in surgery, Simms et al3 conducted 6 replications of a high-fidelity simulation of an actual adverse event. A different subject played the nurse in the key role. None of the subjects were familiar with the original event. This experiment demonstrated that additional, important root causes requiring action can be identified via simulation although they had been missed in the prior RCA process using conventional methods. In another experiment by the same study team, 3 events drawn from a closed claims database were simulated and the participants debriefed.4 Compared with causes noted in the closed claims, more and different root causes were identified. They concluded that “The use of simulation for investigation of adverse surgical outcomes is feasible and identifies causes that may be more amenable to effective systems changes than conventional RCA.”

Lobos et al5 compared findings from simulation versus a traditional investigation approach. They did not use the actual participants in the event but recreated aspects of it with other personnel. Several additional causal factors and recommendations for improvement were identified from the simulation.

In addition to evidence and recommendations from the studies noted previously, other reports call for more use of simulation in investigation of adverse events and suggest how to use simulation for related purposes, for example, training investigators, evaluating recommendations.6 Although RCA is now used commonly in healthcare, the new RCA2 guideline has no mention of re-enactment or simulation of the actual event as a possible tool.2 The event we have described here suggests that simulation should be considered more often for this purpose. One might ask why there are not more publications of such cases or more widespread use of simulation for RCA. We can surmise that there is reluctance to report on cases using RCA for investigating real adverse events, especially where litigation is possible. It is one reason that this case is now being presented, because it happened about 30 years ago.

Because there are additional time and cost associated with simulation, there is likely hesitancy to use it often. Its use must be considered carefully. The criteria could include limitation to events for which possibly key technical or nontechnical causes are unlikely to be revealed without simulation. Perhaps only events with actual or potential serious outcomes should be candidates. Another useful situation may be when only weak corrective actions, eg, education and policy setting, are the ones likely to be identified. Admittedly, it would be difficult to know beforehand which events warrant the time and cost of using simulation.

When using RCA techniques for cases involving medical technology incidents, it is prudent to involve persons with appropriate engineering or equivalent expertise. Biomedical/clinical engineering experts can help explain the function, failure modes, and available information resources related to the involved devices. Some medical devices are inherently dangerous, as is the laser in this case, and such technical and engineering persons can help ensure that simulations are done safely. In some institutions, such expertise may not be available. If so, conducting a re-enactment may not be prudent.

There is not yet a clear set of guidelines for how best to use simulation in this way. For instance, involving the actual participants in the event may not always be advisable. Although they are the people with the most salient knowledge of what happened, their presence together may cause some additional trauma or embarrassment for them. In addition, if there are ill feelings within the team involved in the event, those might surface in ways that are challenging to manage and could create additional discomfort. In the event described here, however, we strongly suspect that the involvement of the actual personnel involved was key to discovering the proximal cause.

If the event is recorded, that can be helpful for later analysis, but the actual participants in the event may not wish to be filmed. Getting their consent is required at a minimum. If they do not wish to be filmed, it may be possible to angle the camera in a way that only captures audio and aspects of the reenactment that do not include identifiable features (this author has done that on a different occasion in more recent years).

Although the focus of this report has been on RCA, simulation and re-enactment are equally applicable to any type of accident investigation where direct and indirect causes, which happen in both linear and nonlinear paths, are not likely to be identified by other means. Root cause analysis endeavors to identify all contributory factors in an event, digging as deeply as possible to seek potential remediation strategies and tactics. Modern safety science has advanced substantially and offers many approaches to conducting such investigations.8

LIMITATIONS

This event happened approximately 30 years ago. Although there does exist some critical documentation, including video of the recreation of the event, some aspects of the event are not known. In addition, there may be some recall bias on the part of the author. Furthermore, a traditional, formal RCA or other investigation technique was not used at the time, and we do not have access to records of the investigation conducted by the risk management organization. In addition, no direct comparison was made between the simulation and an investigation conducted without a simulation. Thus, it is impossible to know with certainty whether the most relevant causes, eg, accidental activation of the footswitch, could have been identified without a recreation/simulation of the event. However, without having the participants recreate what they actually did, using a space and props similar to what were used in the actual case, it is difficult to imagine how the actions involving the footswitch, the placement of the laser wand, and the juxtaposition of the drapes in relation to the injury could have been envisioned.

CONCLUSIONS

There is more to be learned about when and how to best use simulation for optimizing the learning from critical healthcare events. However, knowing of its potential value, it should be a tool in the toolkit of every patient safety professional for whom learning from critical events is a responsibility.

ACKNOWLEDGMENT

The author thanks Dr Mark Bruley for the investigation work he led that provided the material for this commentary and for his review and suggestions on the manuscript.

REFERENCES

1. Sklet S. Methods for accident investigation. Norwegian University of Science and Technology 2002. Available at: http://www.learnfromaccidents.com.gridhosted.co.uk/images/uploads/Norwegian_university_of_science_and_technology_Method_for_accident_investigation.pdf. Accessed April 21, 2020.
2. National_Patient_Safety_Foundation. RCA2: improving root cause analyses and actions to prevent harm. Boston, MA; 2015. Available at: http://www.ihi.org/resources/Pages/Tools/RCA2-Improving-Root-Cause-Analyses-and-Actions-to-Prevent-Harm.aspx. Accessed May 5 2020.
3. Simms ER, Slakey DP, Garstka ME, Tersigni SA, Korndorffer JR. Can simulation improve the traditional method of root cause analysis: a preliminary investigation. Surgery 2012;152(3):489–497.
4. Slakey DP, Simms ER, Rennie KV, Garstka ME, Korndorffer JR Jr. Using simulation to improve root cause analysis of adverse surgical outcomes. Int J Qual Health Care 2014;26(2):144–150.
5. Lobos AT, Ward N, Farion KJ, et al. Simulation-based event analysis improves error discovery and generates improved strategies for error prevention. Simul Healthc 2019;14(4):209–216.
6. Macrae C. Imitating incidents: how simulation can improve safety investigation and learning from adverse events. Simul Healthc 2018;13(4):227–232.
7. Cooper JB, Cullen DJ, Eichhorn JH, Philip JH, Holzman RS. Administrative Guidelines for Response to an Adverse Anesthesia Event. The Risk Management Committee of the Harvard Medical School's Department of Anaesthesia. J Clin Anesth 1993;5:79–84.
8. Oakley JS. Accident Investigation Techniques. 2nd ed. Park Ridge, IL: American Society of Safety Professionals; 2012.
Keywords:

Root cause analysis; accident investigation; patient safety

Copyright © 2020 Society for Simulation in Healthcare